Parameter estimation of one- and two-dimensional heat conduction processes using a minicomputer

Abstract

To estimate unknown parameters in one- and two-dimensional heat conduction processes, a gradient-type algorithm was implemented on a minicomputer. This combination yielded successful results in the experiment for the estimation of constant parameters contained in the system equation (diffusion equation) or in the boundary equation. The results also showed the suitability of, and limitations of, present small computers for high-order (lumped parameter) problems. The relationship between the number of measurement sensors and the convergence of the estimation algorithm was also experimentally studied.